HSZ80 disk failover partially failed

From: <Maurice.Steyvers_at_ICTS.UNIMAAS.NL>
Date: Mon, 22 Jan 2001 10:15:10 +0100

Last week we had the following problem on our Tru64 UNIX DS20E system
connected to a HSZ80 controller.
In one of the RAID0+1 sets a disk failed. As expected the spare disk was
automatically used to replace the failed disk in the set. But unfortunately
somehow UNIX lost connection to the RAID0+1 set regarding the ADVFS-error
that showed up in the sytemlog and the disk-errors in de binary errorlog. A
"cd" command to a directory on the failed disk gave a "permission denied". A
"df" command still showed all filesystems.
The situation could simply be solved by a reboot that made the RAID0+1 set
(and all filesystems on it) available again.

At the time i installed the whole system (about a year ago) and created the
RAID0+1 sets i tested the failover/spare mechanism by unplugging a disk. At
that time everything worked as expected (the spare set was used without UNIX
even noticing anything).

The HSZ80 software version is: HSZ80 ZG95005832 Software V83Z-0, Hardware
E04.

What could be the cause of this ?


______________________________________________

Maurice Steyvers, ICA, Universiteit Maastricht

mailto:maurice.steyvers_at_icts.unimaas.nl
phone :(++ 31 43) 388 23 49
Fax :(++ 31 43) 361 42 87
_______________________________________________
Received on Mon Jan 22 2001 - 09:24:34 NZDT

This archive was generated by hypermail 2.4.0 : Wed Nov 08 2023 - 11:53:41 NZDT